Vertical Profile of C, N, P, K and Radionuclides in Soil Collected from Highland Tea Plantation Areas

Problem statement: Cameron Highlands is a well known agricultural are a in Peninsular Malaysia. Long term usage of fertilizer has caused accumulation of major elemental component in the soil. This accumulation will cause enrichment of th e nutrient in the catchment located at downstream of the river through runoff. Approach: Two tea plantations on the upstream with hilly con dition plantation were selected as the location for monito ri g the accumulation of the major nutrient component N, P and K. C was also determined to esti mate the total organic content in the soil. Natural radionuclides i.e., Ra, Ra and K were also determined and anthropogenic radionucli des Cs were detectable. The samples were measured using El emental Analyzer, Energy Dispersive X-rays Fluorescence (EDXRF) and gamma spectrometer. The da ta set were analyzed using Principle Component Analysis (PCA) and Cluster Analysis (CA) to check the distribution and elemental sources. Results: The trend for all depth profile measurement result s shown monotonically trend through the depth where it shown no observable tren d except for C, N, P and Cs decreasing through the depth. PCA results indicate that there are two sources for plantation A and three sources for plantation B that led to the accumulation of these lements. Three clusters of group element were found for both tea plantation area and the major so urces are from fertilizer, natural occurring and atmospheric natural process. The model for C, N and P was found to be exponentially proportional to the depth with removing mixing layer. Conclusion: The range of concentrations for measured elements shows that the concentrations of elements in tea plantation B are higher than in tea plantati on A. All depth profile gives monotonically trend exce pt for C, N, P and Cs since these elements were added to the soil. C, N and P are decreasing expone ntially with depth. The amount of Cs was found to be detectable for both study locations and it wa s from the fall out of nuclear explosion. Other radionuclides seem to be from natural existence and atmospheric natural process.


INTRODUCTION
Tea plantation in Cameron Highlands is a long life time plantation on high hill slopes that extended to more than 50 years. Long term usage of fertilizer had enriched the cultivated area with major component of fertilizer such as N, P and K in soil. The accumulated fertilizer that is not absorbed by the roots will be easily removed by runoff on the slope surface as well as leaching into the soil. Overused fertilizer will cause nonpoint pollution (Novotny et al., 2010;Guo et al., 2011;Carpenter et al., 1998;Choudhury and Kennedy, 2005). Besides the normal major components chemical fertilizers of potash based and other type of fertilizer also contained appreciable amount of radionuclides (Alam et al., 1997;Chandrajith et al., 2010;Favaro, 2005). The high hill of Cameron Highlands, Malaysia sit on the intermediate igneous rock mountain range (Ministry of Agriculture and Fisheries, 1970). Being igneous in origin the soil is known to contain natural radionuclides of uranium and thorium origins and 40 K (Ademola et al., 2010;Anjos et al., 2005;El-Arabi 2007;Moura et al., 2011). With high altitude of more than 1100 m from sea level, it is suspected that the anthropogenic radionuclide 137 Cs also deposited in the soil (Lettner et al., 2006;McGee et al., 1992). With the hilly location up to 20° gradient steep slope, the runoff is much greater than flat area (Aminuddin et al., 2005).
Previous research at Cameron Highlands shows large amount of sediment accumulation at Sultan Abu Bakar Dam due to soil erosion Riduan et al., 2009). The usage of fertilizer in Malaysia for industrial group including tea plantation has increased each year although without increased in the plantation's area (FAOUN, 2004). Tea plantations occupy 40% fraction of total agriculture area at Cameron Highlands (Aminuddin et al., 2005). Without controlling the usage of fertilizer at tea plantation, the unused fertilizers will increases and also sweep by runoff.

MATERIALS AND METHODS
Sampling: Soil samples were collected at two tea plantations area in Cameron Highlands, Malaysia. The two locations are known as plantation A and plantation B. Seven sampling points were set at plantation A and six sampling point at plantation B and each sampling point location was determined using Global Positioning System (GPS). At each sampling point, ten profile soil samples up to 20 cm were taken as representative sample using hand auger. Each profile samples were sub-sampled at 2 cm slice. Two times drilling were necessary to get a profile up to 20 cm depth. Subsampled of the same layer was mix together. They were oven dried at 60°C until the constant weight achieved. The samples were then ground, homogenized and sieved using 250 µm sieve . Each of the samples was stored in a plastic container, sealed and labeled.
Elemental Analyzer Measurement: C and N were determining using LECO CHNS-932 elemental analyzer. Each sample was weighted at 2.0 mg using micro balance into tin capsules. Five replicate for each sample was measured to maintain the consistent result. Standard sulphamethazine was used to calibrate the instrument. Blank sample was inserted between five replicates to flush the remaining contaminant before starting new sample measurement.
Energy Dispersive X-Ray Fluorescence (EDXRF) measurement: Pellet with 32 mm diameter and 2 mm thickness was prepared by pressing about 2 g of sample using fusion machine at 15 tonne presure. Duplicate samples were prepared and measured using Minipal4 PANalytical bench-top EDXRF and the tube ratings were set to 300 mV, 150 µA with Rh target. Al-thin filter was used for measuring K, Mo filter for U and Th and no filter for P. K α line was used in quantitative of P and K at energy 2.013 keV and 3.312 keV, L α line was used to measured U (13.612 keV) and Th (12.967 keV) with 100 second measuring time as optimize in our laboratory (Abdullah et al., 2011). Standard calibration for P and K was carried out by using series of dilution of K 2 HPO 4 , Standard Reference Material IAEA 312, IAEA 313, IAEA SL-1, IAEA Soil 7 and IAEA SL-1 for U and IAEA RG-Th-1 for Th . Standard addition method was applied to soil samples to eliminate the matrix effect for P and K. The measurement conditions for standard and samples were set to be identical.
Gamma Spectrometer Measurement: About 400 g of samples was placed in plastic container, sealed and leaved for at least 3 weeks for the radionuclides in the samples to reach secular equilibrium state between parent and progenies (Alias et al., 2008). Each sample was counted using HPGe detector for 12 h. The spectrums were analyzed using GammaVision sofware to determine the activity concentration of 226 Ra, 228 Ra, 40 K and 137 Cs.

Statistical analysis:
To confirm the relationship between these elements, Pearson correlation were applied. Principal Component Analysis (PCA) and Cluster Analysis (CA) were used to identify whether the element came from same group perform using statistical analysis software XLSTAT. Figure 1-10 show the average profile concentration of C, N, P, K and radionuclides in soil for plantaions A and B. The Depth profile showed monotonic trend through the depth except for C, N, P and 137 Cs where it trend is decreasing through the depth. The concentration ranges of C, N, P, K, U, Th, 226 Ra, 228 Ra, 40 K and 137 Cs for plantations A and B are tabulated in Table 1.

DISCUSSION
The concentration of C, N and P are decreasing through depth while K, U and Th are not changing very much with depth. The decreasing trend of these elements is due to natural process such as vertical leaching through the depth. Due to the external input from fertilizer on the top of the soil, the upper layer of the soil contains high concentration of N and P. For carbon, the bottom layer undergoes biodegradation and decomposition. The carbon content at the upper layer of soil is higher because of the input from decomposition of organic from leaf, root and other atmospheric deposition. The K show monotonic pattern through the depth, suggested that K is more soluble compare to the others element and the leaching process for K is much higher. Both plantations show different range where concentration at plantation B is much higher than plantation A for C, N, P and K. Although the range is different, the pattern of the depth profile show no significant different.
As for 137 Cs where it came from anthropogenic pollution through atmospheric fallout, it can been seen that (Fig. 10) there were two trends at 2-10 cm and 12-20 cm which suggested they were from nuclear weapon testing in year 1960s and Chernobyl nuclear accident 1986. From Fig. 1-3, at layer 12 and 14 cm, the concentration of C, N and P is slightly higher than at 10 cm layer. The different is occurring due to the sample collection where the sample for profile was drilled twice at the same hole. Two times drilling is for layer 0-10 cm and 12-20 cm will introduce mixing layer at 12 cm and 14 cm. This mixing layer contributed to the observed high concentration at 12 and 14 cm depth.   Table 2 and 3 while dendrogram for plantation A and B are shown in Fig. 11 and 12. It shows that strong correlation between C, N, P and K for plantation A and B except for K at plantation B. Dendrogram show that C, N, P, K and 137 Cs was clustered as C1 group for A and B. U, Th, 226 Ra, 228 Ra and 40 K also clustered as group C2 and C3 as we expect it is from natural origin.
Principal Component Analysis was applied to the whole set data separately for plantation A and plantation B. There are two principal components for plantation A and three principal components for plantation B referring eigenvalues that exceed 1. Plantation A counted for 61.98 and 23.28% of total variation data. Plantation B counted for 64.41%, 18.68 and 10.02% of total variation data respectively with a total cumulative of 85.26 % and 93.11% for plantations A and B. For plantation A, five elements were related to factor 1 in positive direction and three elements in negative direction. C, N, P, K and 137 Cs were related to positive factor and Th, 226 Ra and 228 Ra related to negative factor. U, Th and 40 K are related to factor two.
For plantation B, factor one is related for element C, N, P and 137 Cs. Factor two were strongly for element K, U and 40 K. 226 Ra were strongly related to factor three.
Cluster Analysis (CA) for both plantation (Fig. 11  and 12), cluster C1 group of element consist of C, N, P and 137 Cs was observed for both plantation due to their trend profile that decreasing through the depth and as the elements that added to the soil. The sources of this element suggested it from fertilizer application except for 137 Cs which it from anthropogenic pollution from atmospheric deposition.
U, Th and 40 K was clustered as C2 group for both plantations. 226 Ra was clustered into C3 group and 228 Ra in C2 group for plantation B. 226 Ra and 228 Ra for plantation A was clustered in C3 group. C2 and C3 group was suggested come from natural occurring and atmospheric natural process. The clustering 228 Ra at C3 for Plantation B and C3 at plantation A maybe attributed to the different of altitude, slope, soil type and soil densities of the two study location.
As we observe the trend of C, N and P is decreasing through the depth, we plot an exponential trendline to get the trend within the depth. The correlation coefficient of the exponential is in a good agreement with depth with R 2 more than 0.9 by removing mixing layer.

CONCLUSION
In general the range of elemental concentration at plantation B is higher than plantation A. All depth profile give monotonically trend except for C, N, P and 137 Cs where elements were added to the soil which when applying fertilizer and it may originated from atmospheric process. The concentration of C, N and P are decreasing exponentially with depth. 137 Cs was found at low activity concentration at study locations. Other radionuclide's 226 Ra, 228 Ra and 40 K seem to be from natural existence and atmospheric natural process. Except for 228 Ra, all elements are clustered in their identical respective cluster in both plantation A and B.